Denis Beghin

Genetic differences for nitrogen uptake and nitrogen utilisation efficiencies in winter wheat

Abstract Due to economic and ecological factors, European agricultural practices are likely to go towards extensive systems with lower inputs of nitrogen (N) fertilisers. The objective of this study was to assess varietal differences for N use at two nitrogen levels. A set of 20 winter wheat (Triticum aestivum L.) genotypes was studied over 2 years in northern France on a deep loam soil without (N0) and with 170 kg ha−1 N fertiliser (N+) as ammonium nitrate. Results were consistent on both years as the genotype×year or genotype×year×N level interactions were not significant. The genotype×N level interaction was highly significant except for total N utilisation efficiency (total above-ground dry weight/total above-ground N) and grain N concentration. The genotype×N level interaction for grain yield was mainly due to three contrasting genotypes: Cappelle, a cultivar from the 1940s, had the same yield at N0 and N+; Arche had a high yield at both N levels; and Recital had a high yield with added N and a very low one without N. The number of kernels/ear explained most of the variations of grain yield at N0 (48%) and N+ (80%), and of the interaction (67%). N uptake efficiency (total above-ground N/soil N supply) accounted for 64% of the variation in N use efficiency (grain yield/soil N supply), while at N0 and at N+ it accounted for only 30%. N utilisation efficiency (grain yield/total above-ground N) was then more important at N+ than at N0. Grain N explained most of total plant N variation at both N levels. The interaction for N use efficiency was best explained by the interaction of N uptake (63%). The applications of these results to a breeding programme to create varieties adapted to low-input management systems are discussed.

Genome-wide association mapping of three important traits using bread wheat elite breeding populations

The exponential development of molecular markers enables a more effective study of the genetic architecture of traits of economic importance, like test weight in wheat (Triticum aestivum L.), for which a high value is desired by most end-users. The association mapping (AM) method now allows more precise exploration of the entire genome. AM requires populations with substantial genetic variability of the traits of interest. The breeding lines at the end of a selection cycle, characterized for numerous traits, represent a potentially useful population for AM studies. Using three elite line populations, selected by several breeders and genotyped with about 2,500 Diversity Arrays Technology markers, several associations were identified between these markers and test weight, grain yield and heading date. To minimize spurious associations, we compared the general linear model and mixed linear model (MLM), which adjust for population structure and kinship differently. The MLM model with the kinship matrix was the most efficient. Finally, elite lines from several breeding programs had sufficient genetic variability to allow for the mapping of several chromosomal regions involved in the variation of three important traits.